New AI Act voted and approved by the Parliamentary committee
A first step in the right direction
On May 11th the AI Act’s committee voted and approved a new version of the AI Act, a version that shows that the European Institutions and the AI Act rapporteurs have listened to the concerns of the creative community and to our requests. As you know, the main reason why this key vote was postponed from the end of February to May was the decision of the European Parliament to address generative AI, its potential and its risks.
Now join us in a brief tour of the highlights of the AI Act!
1. The AI Act addresses AI Technologies by dividing them in three categories: those that create unacceptable risk, those that pose a high risk and those that are at a low or minimal risk.
The AI Act now specifically addresses Foundation Models, meaning those AI models built on large amount of data and able to perform different tasks (such as CHAT GPT or Stable Diffusion). The key word when speaking of Foundation Models is “transparency”. Users have to know that a certain content was made by an AI or if they are speaking with a chatbot.
2. A new premise has been added where we can read: “As foundation models are a new and fast-evolving development in the field of artificial intelligence, it is appropriate for the Commission and the AI Office to monitor and periodically asses the legislative and governance framework of such models and in particular of generative AI systems based on such models, which raise significant questions related to the generation of content in breach of Union law, copyright rules, and potential misuse.”
3. Article 28b has been added, “Obligations of the provider of a foundation model”. Point 4 of this article specifically states:
4. Providers of foundation models used in AI systems specifically intended to generate, with varying levels of autonomy, content such as complex text, images, audio, or video (“generative AI”) and providers who specialize a foundation model into a generative AI system, shall in addition
a) comply with the transparency obligations outlined in Article 52 (1),
b) train, and where applicable, design and develop the foundation model in such a way as to ensure adequate safeguards against the generation of content in breach of Union law in line with the generally- acknowledged state of the art, and without prejudice to fundamental rights, including the freedom of expression,
c) without prejudice to national or Union legislation on copyright, document and make publicly available a sufficiently detailed summary of the use of training data protected under copyright law.
As stated by AI Act Rapporteur, Hon. Tudorache, this was a way for the Parliament to bring the copyright protection inside the AI Act. To quote: “What we are saying… is there has to be a documentation and transparency about material that was used by the developer in the training of the model. So that afterwards the holders of those rights… can say hey, hold on, what you used my data, you use my songs, you used my scientific article — well, thank you very much that was protected by law, therefore, you owe me something — or no.”
4. Finally, changes have been made to strengthen the privacy right of European citizens.
We can’t stress enough what an important milestone this text is. When we started our work, back in January, despite having been always met with the highest sympathy and understanding by the European Institutions, we were told that the AI Act was something written in stone, a text designed to protect users and not meant to regulate generative AIs and their training.
Five months later everything has changed.
Does this mean that we are satisfied with this and that our job is done?
Absolutely not!
This is (another) big step in the best direction possible, but the road ahead is long.
At the beginning of June, the plenary of the European Parliament will discuss and vote the AI Act. Our aim is to have some amendments approved in order to make our work and data safer.
After June the AI Act will enter its trilogue phase and much more work will be needed then. During those months, the Parliament will have to negotiate with the European Council on the final form of the text, another occasion for us to intervene and ask for better conditions for the creative community.
While doing all this lobbying work with the European and National Institutions, in the coming weeks we will launch a series of complaints for violations of European and national regulations on the use of personal data and copyright with the Supervisory Authorities operating in the respective European countries.
Of course, now more than ever we need your help and the support of the whole creative community. We are facing months and months of hard work and full commitment to the cause, and we will need all the help we can get to make our voices heard. So, keep sharing the Manifesto, keep talking with your peers and representatives and, most important of all, please consider donating to our cause. We are over 38.000 donated – around 36.000€ once considered Gofundme’s fee – but we need more in order to have all VERA’s activities covered and in order to be able to organize events and meetings to raise awareness among the public opinion.